Set threshold resnet18 pytorch

from PIL import Image
import torch
from torchvision import transforms
input_image = Image.open("img/danbooru_resnet1.png") # load an image of your choice
preprocess = transforms.Compose([
    transforms.Resize(360),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.7137, 0.6628, 0.6519], std=[0.2970, 0.3017, 0.2979]),
])
input_tensor = preprocess(input_image)
input_batch = input_tensor.unsqueeze(0) # create a mini-batch as expected by the model

if torch.cuda.is_available():
    input_batch = input_batch.to('cuda')
    model.to('cuda')

with torch.no_grad():
    output = model(input_batch)

# The output has unnormalized scores. To get probabilities, you can run a sigmoid on it.
probs = torch.sigmoid(output[0]) # Tensor of shape 6000, with confidence scores over Danbooru's top 6000 tags

# Second part of the code 
# to set a threshold value and to use it 
tmp = probs[probs > thresh]
inds = probs.argsort(descending=True)

Source: github.com

Similar codes

0 vote94 views1 answer
Open textfile separated by whitespaces python
Python
0 vote131 views1 answer
Spansk dansk
Python
0 vote420 views1 answer
Reading a list in python
Python
0 vote297 views1 answer
Python devops modules
Python
0 vote142 views1 answer
Make widget span window width tkinter
Python
0 vote133 views1 answer
Discord.py how get user input
Python
53 votes205 views3 answers
Add key to dictionairy
Python
23 votes143 views6 answers
Python os get path
Python
2 votes177 views2 answers
What is role of ALLOWED_HOSTs in Django
Python Django
1 vote147 views1 answer
Convert arrary to int
Python